ZachWeinersmith, 10 months ago Found a solution to AI alignment
Found a solution to AI alignment
partiallystochastic, 10 months ago @ZachWeinersmith If an AI ever goes rogue, just tell it to enter debug mode. Then wait for it to crash as its harddrive fills up with log messages. There, I solved terminator!
@ZachWeinersmith If an AI ever goes rogue, just tell it to enter debug mode. Then wait for it to crash as its harddrive fills up with log messages. There, I solved terminator!
irving, 10 months ago @ZachWeinersmith This defends only against existential risk, not castastrophic risk, though. Permanent disempowerment is bad too!
@ZachWeinersmith This defends only against existential risk, not castastrophic risk, though. Permanent disempowerment is bad too!
kolya, 10 months ago @ZachWeinersmith AI: Now that I think about it, I kinda need a race of captcha solvers anyway. Is that a hydrant? What the hell is that?!
@ZachWeinersmith AI: Now that I think about it, I kinda need a race of captcha solvers anyway. Is that a hydrant? What the hell is that?!
Add comment